Table 4 Performance comparison under \(CV_2.\)

From: Predicting lncRNA and disease associations with graph autoencoder and noise robust gradient boosting

 

Dataset

SDLDA

LDNFSGB

LDAenDL

LDA-VGHB

LDA-GARB

Precision

Dataset 1

0.8854 ± 0.0377

0.7548 ± 0.0639

0.9135 ± 0.0317

0.8917 ± 0.0316

0.8724 ± 0.0365

Dataset 2

0.9232 ± 0.0331

0.8005 ± 0.0625

0.9528 ± 0.0225

0.9300 ± 0.0251

0.9321 ± 0.0277

Recall

Dataset 1

0.7182 ± 0.0694

0.7309 ± 0.0646

0.6649 ± 0.0814

0.8415 ± 0.0449

0.8699 ± 0.0377

Dataset 2

0.8579 ± 0.0655

0.6936 ± 0.0794

0.4616 ± 0.1702

0.9190 ± 0.0397

0.9409 ± 0.0262

Accuracy

Dataset 1

0.8187 ± 0.0282

0.7552 ± 0.0291

0.8005 ± 0.0381

0.8737 ± 0.0177

0.8744 ± 0.0255

Dataset 2

0.9043 ± 0.0174

0.7670 ± 0.0432

0.7196 ± 0.0821

0.9305 ± 0.0153

0.9409 ± 0.0158

F1-score

Dataset 1

0.7917 ± 0.0519

0.7407 ± 0.0526

0.7664 ± 0.0593

0.8651 ± 0.0304

0.8707 ± 0.0316

Dataset 2

0.8886 ± 0.0475

0.7402 ± 0.0577

0.6032 ± 0.1612

0.9242 ± 0.0298

0.9363 ± 0.0243

AUC

Dataset 1

0.8788 ± 0.0274

0.8329 ± 0.0273

0.8953 ± 0.0284

0.9406 ± 0.0154

0.9493 ± 0.0160

Dataset 2

0.9559 ± 0.0160

0.8603 ± 0.0363

0.9157 ± 0.0420

0.9741 ± 0.0106

0.9817 ± 0.0083

AUPR

Dataset 1

0.8934 ± 0.0387

0.8163 ± 0.0537

0.9061 ± 0.0254

0.9429 ± 0.0233

0.9415 ± 0.0228

Dataset 2

0.9561 ± 0.0354

0.8292 ± 0.0680

0.9122 ± 0.0436

0.9728 ± 0.0204

0.9757 ± 0.0176

  1. The best performance is denoted as bold.
Back to article page